DigitalOcean recently introduced its GenAI platform, which allows developers to build and deploy AI agents quickly and easily without experience.
DigitalOcean is a niche cloud infrastructure player targeting smaller companies, especially developers within them. With a serverless infrastructure, the cloud provider enables this target group to easily design, build and quickly put applications into production.
DigitalOcean’s new GenAI platform allows customers to design, build and deploy AI agents. The platform targets explicitly anyone who wants to develop such AI agents, including developers with little or no knowledge of AI. The platform acts as a kind of blueprint for building these agents.
Functionality
Users can use an AI chatbot to describe exactly what they want and choose from a library of powerful foundation models that target specific tasks, such as document analysis, image generation, or semantic search. The selected LLMs are then paired with proprietary or public datasets to build and deploy the AI agents.
Function calling connects the created AI agents to databases or APIs so that users always have the most up-to-date and accurate data. In addition, security restrictions are built to ensure the AI solutions’ reliability and safety.
These security measures are designed to minimize the risk of erroneous, unreliable or unwanted responses. In addition, private endpoints are supported.
Other platform features include the “framework agnostic” structure, no-code functionality and the ability to use the company’s popular “GPU Droplets” – VMs supported by Nvidia H100 GPUs.
Tip: DigitalOcean acquires Cloudways for $350 million
Deployment and roadmap
According to DigitalOcean, the GenAI platform is especially suited for AI agents that leverage specific context-, language- and customer-specific data. This results in simple and accessible workflows, ideal for users without AI experience who still want to work with AI agents.
The cloud infrastructure specialist sees particular opportunities for applications in customer support, manufacturing and automation, and education and training purposes.
Shortly, DigitalOcean plans to expand the GenAI platform, which is now available to everyone after a beta phase, to include features such as supporting URLs as data sources, agent evaluations, CI/CD pipelines, automatic indexing of knowledge resources and fine-tuning of LLMs.